Eight trained male cyclists (age 20-33 yr) completed four 3-h bouts of cycling at 60% peak VO2 in the heat (33 degrees C) drinking either water (W), 5% glucose (G), 5% glucose polymer (GP), or 3.2% glucose polymer + 1.8% fructose (GP/F) at a rate of 350 ml every 20 min (3.15 l total volume). Similar changes in heart rate, sweat rate, rectal and mean skin temperatures, and plasma [Na+], [K+], and osmolality were observed during all trials. Mean changes in plasma volume, although not significantly different between trials, were lowest for the GP/F drink (-2.6%) and greatest for the G (-8.1%) drink. Plasma volume decreased (P less than 0.05) below pre-exercise control values during the W, G, and GP trails but was maintained at control values during the GP/F trials. In contrast to water ingestion, G, GP, and GP/F ingestion maintained plasma glucose and respiratory exchange ratios throughout the 3-h exercise bouts. Gastric residual volume (GRV) obtained at the end of exercise was similar for the W, GP, and GP/F trials. The G trials yielded greater (P less than 0.05) GRV than W trials. For all drinks ingested, over 90% of the 3.15 l consumed was emptied from the stomach during the 3-h exercise bouts. At a mean sweat rate of 1.2 l.h-1, cyclists replaced 73% of fluid lost and experienced only a 1.6% loss in body weight. This study demonstrates that, during prolonged (3-h) cycling exercise in the heat, large volumes of W and 5% carbohydrate can be emptied from the stomach to help minimize the effects of dehydration.